Inline AudioFlinger::initCheck and remove unnecessary lock.
Remove redundant check of mAudioHwDevs.size().
No need to lock mHardwareLock for each device separately
during initialization.
Use size_t not int to loop through Vector, since size() returns size_t.
Add missing hardware lock for get_mic_mute() and get_input_buffer_size().
Add comments.
Change-Id: Iafae78ef78bbf65f703d99fcc27c2f4ff221aedc
We can remove mExiting and use Thread::exitPending() instead.
The local sp<> on "this" in exit() is not needed, since the caller must
also hold an sp<> in order to be calling us. (Unless it was using a raw
pointer, but that would be dangerous for other reasons.)
Add comment explaining the mLock in exit().
Change-Id: I319e5107533a1a7cdbd13c292685f3e2be60f6c4
Put IAudioFlinger methods in binder opcode order.
Move hardware call state closer to where it is used.
getMode() and btNrecIsOff() are private.
Change-Id: Ie50340b396c39c763f2b155cbc08da8a0d0f2424
Fix race conditions when setting master volume, master mute, stream
volume, stream mute for a playback thread, and when reading stream
volume of a playback thread. Lock order is AudioFlinger, then thread.
Rename streamVolumeInternal to streamVolume_l, comment, and use it to
implement streamVolume().
Code size reduction:
- Remove dead code: AudioFlinger::PlaybackThread::masterVolume, masterMute, streamMute.
- Change return type of non-binder methods that always succeed from status_t to void.
- Remove virtual from volume and mute methods that don't need it.
This change saves 228 bytes but decreases performance of binder operations
due to the added locks.
Change-Id: Iac75abc1f54784873a667d1981b2e08f8f31e5c9
stream_type_t is used by AudioFlinger class, so it should be declared there.
This way we don't have to peek into PlaybackThread to get the declaration.
Change-Id: Ie08bab1604699214d1e8df2d48d3fbfbbc436e96
This avoids possible confusion with thread's type().
Also remove redundant cast "(audio_stream_type_t)".
Change-Id: I320b9177b6c267a102d215f002228bcf988c437a
Other:
- add a comment to nextUniqueId
- made ThreadBase::mId const, since it is only assigned in constructor.
Change-Id: I4e8b7bec4e45badcde6274d574b8a9aabd046837
This is just documentation, as C++ method const-ness doesn't mean anything
for a binder API. Instead, here const means "no side effects".
Change-Id: Iaa9cd2fe477db10ae9a40cac4f79f0faa9b4e5e6
It turns out to be just a comment, as all except AudioMixer are RefBase.
There are only a few performance-sensitive cases where it's worth thinking
about whether you need a virtual destructor, and the headache usually
outweighs the benefit.
Change-Id: I716292f9556ec17c29ce8c76ac8ae602cb496533
In constructors, initialize member fields in the initialization list
rather than constructor body where possible. This allows more fields
to be const, provided they are never modified.
Also initialize POD fields in constructor, unless it's obvious they
don't need to be initialized. In that case, put a comment instead.
Remove explicit clear() in destructors on fields that are now const.
Give AudioSessionRef a default constructor, so it's immutable fields can
be marked const.
Add comment about ~TrackBase() trick.
Initialize fields in declaration order to make it easier to confirm that
all fields are set.
Move initialization of mHardwareStatus from onFirstRef() to constructor.
Use NULL not 0 to initialize raw pointers in initialization list.
Rename field mClient to mAudioFlingerClient, and getter from client()
to audioFlingerClient().
Change-Id: Ib36cf6ed32f3cd19003f40a5d84046eb4c122052
EffectModule::addHandle and Client::heap() were declared incorrectly.
As a parameter, an sp<> should be & for efficiency, and for input
parameters it should also be const to protect the caller's value.
But as a return value, an sp<> should have neither const or &. The "e"
in "return e;" might be located on the stack, and if there is "&" then
the caller would see the address of a variable which no longer exists.
Also, an & would make it hard to do "return 0;".
A "const" without & is meaningless in the return type.
(In this particular case, the "e" is a member field, so it was safe.)
Change-Id: I3df5f294214eb15a9d4d596c6d5ef29de97b5c27
Use type_t instead of int for thread types.
Initialize ThreadBase::mType in constructor and make it const.
Change-Id: I43d141388b9639e4783c30b97dbda5688bf7555f
Was int or uint32_t.
When AudioFlinger::format can't determine the correct format,
return INVALID rather than DEFAULT.
Init mFormat to INVALID rather than DEFAULT in the constructor.
Subclass constructors will set mFormat to the correct value.
Change-Id: I9b62640aa107d24d2d27925f5563d0d7407d1b73
This problem due to the way audio buffers are mixed when
low power mode is active was addressed by commits 19ddf0eb
and 8a04fe03 but only partially. As a matter of fact, when more
than one audio track is playing, the problem is still present.
This is most noticeable when playing music with screen off
and a notification or navigation instruction is played: in this case,
the music or notification is likely to skip.
The fix consists in declaring the mixer ready if all active tracks
are ready. Previous behavior was to declare ready if at least one track was
ready. To avoid that one application failing to fill the track buffer blocks other
tracks indefinitely, this condition is respected only if the mixer was ready
in the previous round.
Issue 5799167.
Change-Id: Iabd4ca08d3d45f563d9824c8a03c2c68a43ae179
A bad parameter to AudioFlinger::createTrack could cause mediaserver to crash.
Other AudioFlinger stream type cleanup:
- Simplify range check for audio_stream_type_t
- Add comment about mStreamTypes array initialization.
Change-Id: Ia33aa1cce0fdd694b08d9288816ffc097a9543d0
